Urban storm flood simulation using improved SWMM based on K‐means clustering of parameter samples

Abstract

Abstract To address the two problems of unclear delineation of sub‐catchment and complicated and cumbersome parameter rate determination in the Storm Water Management Model (SWMM), this study proposes a rapid construction method of SWMM based on the principle of single urban functional area combined with K‐means clustering algorithm, The research area is the southern part of Jinshui District, Zhengzhou City. The Hydrological Response Unit (HRU) contains only a single urban functional area, divided by combining the natural and social attributes of the urban surface. Calibrated uncertain parameters from 76 papers were selected as samples, and the K‐means clustering algorithm was used to cluster and calculate the parameter values, to improve the SWMM model, selecting three typical rainfall runoff processes for validation application. The results show that simulated runoff is consistent with measured runoff trends, with the NSE and R2 value scores of the flow processes of the three floods above 0.86 and the, locations and numbers of flooded nodes are consistent with the actual research. This provides a new idea and technical support for the construction of urban flood models in flood prevention and mitigation. The relevant results can provide scientific decision‐making reference for urban flood forecasting and warning.

Keywords

• division of HRU
• flooding nodes
• K‐means cluster algorithm parameter determination
• SWMM
• urban storm flood model